Potent, Selective, Water Soluble, Brain-Permeable EP2 Receptor Antagonist for Use in Central Nervous System Disease Models

Article abstract of DOI:10.1021/acs.jmedchem.9b01218

Activation of prostanoid EP2 receptor exacerbates neuroinflammatory and neurodegenerative pathology in central nervous system diseases such as epilepsy, Alzheimer's disease, and cerebral aneurysms. A selective and brain-permeable EP2 antagonist will be useful to attenuate the inflammatory consequences of EP2 activation and to reduce the severity of these chronic diseases. We recently developed a brain-permeable EP2 antagonist 1 (TG6-10-1), which displayed anti-inflammatory and neuroprotective actions in rodent models of status epilepticus. However, this compound exhibited moderate selectivity to EP2, a short plasma half-life in rodents (1.7 h) and low aqueous solubility (27 μM), limiting its use in animal models of chronic disease. With lead-optimization studies, we have developed several novel EP2 antagonists with improved water solubility, brain penetration, high EP2 potency, and selectivity. These novel inhibitors suppress inflammatory gene expression induced by EP2 receptor activation in a microglial cell line, reinforcing the use of EP2 antagonists as anti-inflammatory agents.

Full text of DOI:10.1021/acs.jmedchem.9b01218

Subscriber access provided by UZH Hauptbibliothek / Zentralbibliothek Zuerich
Article
Potent, selective, water soluble, brain-permeable EP2 receptor
antagonist for use in central nervous system disease models
Radhika Amaradhi, Avijit Banik, Shabber Mohammed, Vidyavathi Patro, Asheebo
Rojas, Wenyi Wang, Damoder Reddy Motati, Raymond Dingledine, and Thota Ganesh
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b01218 • Publication Date (Web): 06 Jan 2020
Downloaded from pubs.acs.org on January 7, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Potent, selective, water soluble, brain-permeable
EP2 receptor antagonist for use in central nervous
system disease models
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Radhika Amaradhi, Avijit Banik, Shabber Mohammed, Vidyavathi Patro, Asheebo Rojas, Wenyi
Wang, Damoder Reddy Motati, Ray Dingledine and Thota Ganesh*
Department of Pharmacology and Chemical Biology, Emory University School of Medicine,
1
510 Clifton Rd; Atlanta, GA, 30322, United States of America.
KEYWORDS: Neuroinflammation, PGE
2
-receptors, Schild K , competitive antagonist, lead
B
optimization, SAR, blood-brain barrier, aqueous solubility.
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT: Activation of prostanoid EP2 receptor exacerbates neuroinflammatory and
neurodegenerative pathology in central nervous system diseases such as epilepsy, Alzheimer’s
disease and cerebral aneurysms. A selective and brain-permeable EP2 antagonist will be useful to
attenuate the inflammatory consequences of EP2 activation and to reduce the severity of these
chronic diseases. We recently developed a brain-permeable EP2 antagonist 1 (TG6-10-1), which
displayed anti-inflammatory and neuroprotective actions in rodent models of status epilepticus.
However, this compound exhibited moderate selectivity to EP2, a short plasma half-life in rodents
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(1.7 h) and low aqueous solubility (27 µM), limiting its use in animal models of chronic disease.
With lead optimization studies, we have developed several novel EP2 antagonists with improved
water solubility, brain-penetration, high EP2 potency and selectivity. These novel inhibitors
suppress inflammatory gene expression induced by EP2 receptor activation in a microglial cell
line, reinforcing the use of EP2 antagonists as anti-inflammatory agents.
2
ACS Paragon Plus Environment

Page 3 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
Neuroinflammation has emerged as a key feature that exacerbates chronic neurodegenerative
pathology in several central nervous system diseases including epilepsy, Alzheimer’s disease (AD)
and Parkinson’s disease.1-4 Neuroinflammation is also an early event that plays a deleterious role
in status epilepticus and traumatic brain injury.5, 6 Therefore, anti-inflammatory agents should offer
a benefit to patients with these central nervous system (CNS) diseases. Production of inflammatory
cytokines (interleukins) and chemokines, activation of glia, and induction of cyclooxygenase-2 are
the early features of neuroinflammation.7, 8 Thus, there is an excellent opportunity for therapeutic
discovery targeting the mediators of neuroinflammatory signaling.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
A significant body of evidence suggests that neuroinflammation precedes the appearance of
amyloid-β plaques and neurofibrillary tangles, which themselves precede clinical signs of AD in
9
animal models and patients. Among the neuroinflammatory inducers, COX-2 has been very well
studied thus far in AD patients by using small molecule inhibitors of COX-2.10-12 Although
retrospective epidemiological studies using non-selective COX-2 inhibitors suggested that chronic
use of COX-2 inhibitors might prevent or delay the onset of Alzheimer’s disease in patients,
prospective studies concluded that selective COX-2 drugs do not provide a clear benefit for
1
3
Alzheimer’s patients. Reconciliation of the epidemiologic studies with prospective clinical trial
results might be found in the clinical trial design, in which the treatment of patients was initiated
when the disease is already at advanced stages in terms of classical pathological markers β-amyloid
plaques and neurofibrillary tangles, or insufficient duration of treatment. The prevailing hypothesis
emerging from clinical trials of a COX-2 drug is that treatment of patients starting at the prodromal
stage of disease could offer benefit, but not treatment starting at late stages of the disease. Testing
this proof of concept requires administration of a drug for several months (in animals) to years (in
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
human) before clinical dementia symptoms are diagnosed. Regrettably, the COX-2 inhibitors
cause adverse cardiovascular effects upon chronic use, by blunting PGI
synthesis.14, 15 Two
2
selective COX-2 inhibitors rofecoxib (Vioxx) and valdecoxib (Bextra) were withdrawn from the
USA market, but a less selective COX-2 inhibitor, celecoxib (Celebrex), is in the clinic with black-
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
6
box warning for thrombotic cardiotoxicity. Thus, future anti-inflammatory therapy within the
COX-2 pathway should be targeted through a specific proinflammatory prostanoid synthase or a
receptor downstream of COX-2, rather than blocking the entire COX-2 signaling pathways.17, 18
COX-2 catalyzes the synthesis of five prostanoids, PGD
2
, PGE
2
, PGF
2
, PGI
2
and TXA . These
2
prostanoids activate nine receptors, DP1, DP2, EP1, EP2, EP3, EP4, FP, IP and TP respectively.
Each of these receptors can play protective as well as deleterious roles in CNS and peripheral
pathophysiologies.17, 18 The EP2 receptor has emerged as a predominantly proinflammatory
receptor that mediates much of the COX-2 driven inflammatory consequences.19, 20 When activated
by PGE , EP2 receptors stimulate adenylate cyclase resulting in elevation of cytoplasmic cAMP
2
concentration, which initiates downstream events via cell signaling pathways mediated by protein
kinase A or exchange protein activated by cAMP (Epac).17, 18
Genetic ablation of the EP2 receptor reduces oxidative damage and amyloid- burden in a mouse
21
model of AD. EP2 deletion also attenuates neurotoxicity and reduces α-synuclein aggregation in
22
a mouse model of PD. Moreover, EP2 deletion improves motor strength and survival of the ALS
2
3
mouse. Furthermore, PGE signaling via EP2 receptor increases injury in models of cerebral
2
2
4
ischemia, and mice lacking EP2 receptors have shown less cerebral oxidative damage produced
25
by the activation of innate immunity. EP2 activation promotes inflammation and neurotoxicity
in models of status epilepticus induced by pilocarpine and diisopropylfluorophosphate (DFP).19,
4
ACS Paragon Plus Environment

Page 5 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
26, 27
In vitro, microglia cultures from mice lacking EP2 have shown enhanced amyloid-
2
8
phagocytosis and are less sensitive to amyloid- induced neurotoxicity. Although these results
strongly support the notion that pharmacological inhibition of EP2 should be explored in
preclinical and clinical setting, the currently available EP2 antagonists have sub-optimal drug-like
features such as low in vivo metabolic stability, in vivo plasma half-life and aqueous solubility
for chronic dosing and proof-of-concept testing in preclinical AD models.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
29
30, 31
32
Initially Pfizer, our laboratory
and recently Amgen identified four distinct classes of EP2
antagonists (Figure 1A). Only the Amgen compounds seems to be useful for chronic dosing, but
the derivatives developed in Pfizer and in our laboratory displayed too low liver metabolic
stability, plasma half-life, or brain permeability to be useful in chronic CNS disease models such
as Alzheimer’s disease. We initially identified a cinnamic amide compound TG4-155 (Figure 1A)
3
0
from high-throughput screening as a potent EP2 antagonist. Structure activity relationship study
of this compound resulted in the first generation brain-permeable lead compound 1 (TG6-10-1,
Figure 1B), which displays brain-to-plasma ratio of 1.7, but only 10-fold selectivity for EP2 over
3
3
DP1, low aqueous solubility (27 μM) and moderate plasma half-life (1.7 h). Thus, developing a
selective, aqueous-soluble compound with enhanced plasma half-life while maintaining potency
and brain permeability is the main goal of our project. In the present study, we report further lead-
optimization studies that led to the synthesis of several second-generation analogs and show that
improvements are made in terms of selectivity, aqueous solubility and in vivo plasma half-life.
Moreover, we developed a compound 20o.HCl (named TG11-77.HCl), which is water-soluble
(2.52 mM), yet crosses the blood-brain-barrier with brain-to-plasma ratio of 0.4. Furthermore,
several novel compounds in this class, including 20o.HCl exhibited anti-inflammatory activity in
EP2-expressing microglia-derived BV2 cell cultures in vitro. Therefore, compound 20o.HCl and
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
other derivatives in this class are now suitable candidates for dosing into - disease models such as
AD.
RESULTS AND DISCUSSION
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Lead optimization studies to improve DMPK properties.
Figure 1. A. First generation EP2 antagonist hits reported in the literature. B) Lead optimization
pathway identifies a brain-permeable EP2 antagonist candidate 20o.HCl (TG11-77.HCl) from
previous lead EP2 antagonist 1 (TG6-10-1)
6
ACS Paragon Plus Environment

Page 7 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Previous SAR studies around compound 1 (TG6-10-1, Figure 1B) indicated that the acrylamide
moiety is not essential for EP2 bioactivity. Thus, we developed second-generation analogs with an
amide linker for SAR study. Compound 2a and 3 (Figure 1B) are examples of the earlier leads.
These two compounds displayed high EP2 potency and selectivity and improved aqueous-
solubility. However, compound 2a displayed poor DMPK properties, for example < 2 minutes
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
4
half-life in mouse liver microsomes. Although compound 3 displayed excellent stability in mouse
liver microsomes (> 60 min) and in vivo plasma half-life in mice (10.5 h), it showed very poor
35
brain-to-plasma ratio of 0.02. Therefore, it was not useful for proof-of concept studies in CNS
disease models. The overall goal was to develop a compound with requisite brain penetration and
plasma half-life. In the past, we have made several derivatives replacing the right-side ring
(morpholine in 2a, tetrazole in 3) and learned that this ring region can be modified without
substantial loss of potency at EP2 receptor. We also made limited derivatives for example, polar
3
4
imidazole ring replacing the indole on the left-side ring to increase the aqueous solubility. We
have not made modifications on the middle phenyl ring for SAR study in the past. Therefore, we
expanded our lead optimization studies with structural modification in these areas as detailed
below.
Chemical synthesis
Synthesis of novel derivatives.
To explore the middle phenyl ring for structural modification and SAR studies, novel compounds
2b-c have been synthesized in a single step (Scheme 1) using commercially available precursors
34
(
1b and 1c) by following the synthetic method reported for 2a. However, several other precursors
(e.g. 6a-g, 10a-c, 13a-c) needed for synthesis of novel derivatives were not commercially
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
available, therefore we have synthesized here as shown in Scheme 2. Commercially available
boronic acid derivatives (4a-g) have been coupled to 6-bromonicotinic acid (5) in the presence of
tetrakis(triphenylphosphine)palladium (0) catalyst and a base to provide intermediates 6a-g, which
were then coupled to 2-(2-methyl-1H-indol-3-yl)ethan-1-amine (7) in the presence of 1-ethyl-3-
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(3-dimethylaminopropyl)carbodiimide.hydrochloride (EDCI.HCl) and (dimethylamino)pyridine
(DMAP) to provide final compounds 8a-g. Analogously, boronic acid derivatives 4a-c on
palladium catalyzed coupling with 5-bromopicolinic acid (9) resulted in intermediate derivatives
1
4
0a-c, which were further coupled to 7 to provide compounds 11a-c. Moreover, the boronic acid
c was coupled to 2-chloropyrimidine-5-carboxylic acid (12a) or 4-bromo-3-fluorobenzoic acid
(12b) or 4-bromo-benzoic acid (12c) in the presence of palladium (II) catalyst and a base to result
in intermediates 13a-c, which were similarly coupled with amine 7 to furnish final compounds
1
4a-c.
Scheme 1. Synthesis of middle ring modified derivatives 2b and 2ca
8
ACS Paragon Plus Environment

Page 9 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Scheme 2. Synthesis of middle ring modified EP2 antagonist derivativesa
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Furthermore, we synthesized a number of derivatives with ‘NH’ as a linking moiety between the
middle ring and the last ring of the molecule. As shown in Scheme 3, the amine 7 was coupled to
2
-aminopyrimidine-5-carboxylic acid (15) in presence of coupling reagent EDCI.HCl to provide
intermediate amine 16, which was further coupled to 2-bromopyridines 17a-o in the presence of
tris(dibenzylideneacetone)dipalladium (0) and a base to provide final products 20a-o.
Analogously, intermediate 16 was coupled to commercially available 4-bromopyridine (18) and
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
-bromopyridines 19a-b to provide 20p and 20q-r respectively as final products in good yields.
Scheme 3. Synthesis of middle and right side ring modified EP2 antagonist derivativesa
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Maintaining the middle ring as pyrimidine, and the last ring as 4,6-dimethyl pyridine (see next
section for SAR discussion), we further synthesized novel compounds with substitutions on the
left side indole ring. As shown in Scheme 4, ethyl-2-aminopyrimidine-5-carboxylate (21) on
reaction with 17o in presence of palladium (II) acetate gave intermediate 22, which on hydrolysis
afforded acid precursor 23. Requisite amines 7a-e (SI Figure 1) were synthesized using the known
methods, 36, 37 and they were individually coupled with 23 in the presence of DMAP and EDCI.HCl
o
at 50 C to furnish the final products 24a-e (Scheme 4). Similarly, compound 24f was synthesized
3
8
using acid 23 and 2-trifluroindole-3-ethanolamine (7f), which was intern prepared form 2-vinyl-
pyrrolidin-2-one (SI Figure 1). Moreover, the compound 2-(2-methylpyrazolo[1,5-a]pyridin-3-
yl)ethan-1-amine (7g) has been synthesized starting from amino-pyridinium iodide (SI Figure 1)
3
9
. Compound 7g, and the commercially available 2-amino-1-(2-methyl-1H-indol-3-yl)ethan-1-
10
ACS Paragon Plus Environment

Page 11 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
one (7h) and 2-amino-1-(2-methyl-1H-indol-3-yl)ethan-1-ol (7i) were coupled to the acid
precursor 23 in the presence of EDCI.HCl coupling reagents to provide final products 24g-i,
respectively in moderate yields. All the intermediates and final products were characterized by
complementary methods of NMR and LCMS. The compounds having purity > 95% were subjected
to testing in bioassays as described below.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 4. Synthesis of left side indole and ethylene linker modified analogsa
SAR analysis of novel derivatives
Incorporation of nitrogen into the structure of small molecules often enhances aqueous solubility,
intestinal permeability and other DMPK properties.40 Therefore, we envisioned hitherto
unexplored middle phenyl ring (see ring assignments in Figure 1B) for structural modification
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
with nitrogen heterocyclic rings for SAR studies. Recently, several groups have analyzed the
physicochemical properties of CNS candidates and drugs to propose a multi-parameter
optimization (MPO) approach that can be used to design novel chemical agents with CNS desirable
properties.41-43 Based on this MPO approach, Wager et al. proposed six physicochemical properties
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(MW, cLogP, cLogD, HBD, TPSA and pKa, Table 1) that cumulatively guide CNS activity and
4
3
permeability of molecules. Our previous lead compounds 2a and 3 attained desirable MPO score
(>4 out of 6 maximum), however, compound 3 was found to be a CNS impermeable compound
with in vivo brain-to-plasma ratio 0.02, whereas compound 2a was too short lived in vivo to
determine the brain permeability. We also determined that 3 is not a substrate of efflux pumps, but
-6
it displayed low cell permeability Paap (A-B) of 0.41 x 10 cm/sec in MDR1 expressed MDCK-
cell monolayers with efflux ratio 0.3 (Table 7). Examination of its properties (Table 1) indicate
that further optimization of the total polar surface area and hydrogen bond donor (HBD) properties
4
4
might improve CNS permeability. Moreover, compound 3 has a tetrazole ring, which is known
to act as a surrogate for carboxylic acid, which could limit a passive brain permeability.45
Therefore, we used compound 2a as a starting point for the current lead optimization studies.
Table 1. CNS desirable properties suggested by MPO approach and calculated biophysical
properties of selected EP2 antagonists including earlier leads 2a, 3 and current lead 20o.HCl
a
Property
Property range for
CNS drugs and
candidates
2a
3
8c
8g
14a
20o.HCl
MW/FW
cLogP
305-360
2.3-3.3
363
346
385
415
386
436
3.25
2.64 4.02
1.04 4.02
3.86
3.49
3.58
cLogD
1.7-2.2
3.25
3.86
3.49
3.57
(pH 7.4)
12
ACS Paragon Plus Environment

Page 13 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
pKa
8.4
1
0.06
2
-0.65 2.28
2.83
2
0.59
2
4.5
3
HBD
3
2
PSA (Ǻ2)
44.8-51.2
57.36
96.4 67.0
5.0 3.8
76.2
79.9
95
MPO
score
desired MPO score
> 4
4.7
3.7
4.3
3.6
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
aChemAxon software was used to calculate the values for compounds 2a, 3 and 20o listed in
Table 1
We previously reported compound 2a with good EP2 potency and selectivity which showed
3
4
moderate aqueous solubility. It has a morpholine ring on the right, phenyl ring in the middle, and
3
-indole moiety on the left side (Figure 1B). To learn whether the middle phenyl ring can be
replaced with other rings, we synthesized and tested compound 2b with a nicotinic acid ring.
Compound 2b displayed 4-fold less EP2 potency than 2a, but it showed 3-fold increased aqueous
solubility reinforcing the notion that nitrogen in the ring will enhance the aqueous solubility.40
Keeping the nitrogen (nicotinic) in the middle ring, we substituted right side morpholine ring with
piperidine ring (2c). This resulted in 6.4-fold loss of potency in comparison to 2a. However,
replacement of morpholine with a phenyl ring on the right side resulted in compound 8a, which
interestingly displayed nearly equal EP2 potency to 2a, but low aqueous solubility (Table 2).
Incorporation of a methoxy group either at 3-position (8b), or methoxy or hydroxy at 2-position
(8c-8d) on the phenyl ring improved the potency by 2-fold (~10 nM, Table 2) in comparison to
2
a. Whereas, fluorine (8e) at 2-position reduced the potency by 2-fold in comparison to 2a, and 4-
fold in comparison to 2-methoxy derivatives 8c. Interestingly, incorporation of 4-acetamido group
(8f), or two methoxy groups at 3- and 5-positions (8g) further enhanced the potency with K values
.4 and 2.9 nM respectively. To see if changing the location of nitrogen atom in the middle ring
B
4
from meta-to-ortho (see o-m positional assignments in Scheme 2), we synthesized derivatives 11a-
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
c. Compound 11a was 4-fold less potent than equivalent derivative 8b. Similarly, 11b was 11-fold
less potent than equivalent derivative 8g, and compound 11c was 3-fold less potent than equivalent
compound 8f. These observations suggest that nitrogen atom is preferred at m-position than at o-
position of the middle ring (meaning nicotinic ring is more favorable than picolinic ring) for
optimal potency. Furthermore, we synthesized a pyrimidine derivative 14a and it showed 2.9-fold
less potency than the corresponding nicotinic ring compound 8c, but it has the same potency as
initial lead 2a in the current study. Nonetheless, nicotinic middle ring maintains high EP2 Schild
potency (< 50 nM), except for 2b-c. It is important to emphasize here that derivatives 8a-g and
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
4a displayed enhanced solubility in simulated gastric fluid at pH 2.0, whereas compounds 11a-c
have not showed solubility enhancement in the simulated gastric fluid (Table 2), suggesting
nicotinic ring derivatives will have better aqueous solubility than picolinic ring derivatives. It is
worth to indicate that the nitrogen in the middle ring is not absolutely essential for EP2 potency,
because compound with phenyl ring, 14c has also showed high EP2 potency (cf. 8c), but a fluorine-
substitution on the middle phenyl ring drastically reduced the potency by 25-fold (cf. 14b vs. 14c
or 8c) hinting that additional substitutions on the middle ring might not be tolerated for potency.
Table 2. Middle ring modified EP2 antagonists. Potency and aqueous solubilitya
Entry
EP2
Aqueousb
Solubility
in SGF
c
K
B
Solubility
(nM)
(µM)
(µM)
R =
2
a
28.6
118
70
ND
2
b
206
ND
2
c
185
ND
25
ND
8
a
23.2
>100
8
b
9.5
13
>100
14
ACS Paragon Plus Environment

Page 15 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
8
c
10
28
8
>100
50c
8
d
11.5
8
e
48.4
4.4
22
25
12
>100
>100
>100
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
8f
8
g
2.9
11a
11b
48.6
33.6
14
10
14
10
1
1c
12.8
29.6
10
71
10
1
4a
>100
250
7.7
10
5
10d
ND
14b
4c
1
aThe potency of the compounds is presented in the form of Schild K
using the formula log (dr-1) = log X –log K
by the antagonist compound, X is antagonist concentration 1 μM. K
concentration required to produce a 2-fold rightward shift of PGE values are average of
measurements run in duplicate. The solubility of the compounds is measured in PBS buffer (pH
values, which are calculated
B
B
B
, where dr (dose ratio) = fold shift in EC50 of PGE
value indicates a
2
B
B
2
EC50. K
B
b
2
7
.4) with 1% DMSO by nephelometry.4 The numbers are derived by measuring solubility in
6 c
simulated gastric fluid (SGF) at pH 2.0 with 1% DMSO by nephelometry.
Although the requisite EP2 potency is obtained for several analogs shown in Table 2, only
compound 8c has displayed good selectivity (>100-fold) against other receptors (see Table 6 and
next section for selectivity discussion). Moreover, when representative compounds (8c, 8d, 8g,
1
4a) from the Table 2 were tested for stability in mouse liver microsomes, they displayed very
short half-life (Table 5). Therefore, we envisioned synthesizing additional novel derivatives by
keeping the middle ring as pyrimidine and modifying the third ring (Scheme 3). The other key
difference between these novel set of derivatives shown in Table 3 in comparison to the ones
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
shown in Table 2 is the NH functional group, which links the third ring and the middle pyrimidine
ring. We envisioned that the NH group should enhance the molecular flexibility and potentially
allow us to make anionic salts. In the process, we first synthesized and tested 2-pyridyl derivative
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
0a. Interestingly, this compound displayed high EP2 potency with K
B
= 10.7 nM. Addition of
= < 10 nM).
methyl group (20b) or a fluorine (20c) at 4-position maintained high potency (K
B
Moving the methyl group from 4- to 6- position also maintained a high potency (cf. 20e vs. 20b),
but similarly moving the fluorine atom to 6- position recued the potency by 16-fold (cf. 20f vs.
2
0c). A methoxy group at 6- position also maintained high EP2 potency (20h), whereas cyano-
group reduced the potency by 23-fold (20g). The acetyl group on pyridine ring whether it is at 6-
,5- or 4- position (20j, 20l and 20k) reduced the potency by 4-, 20-and 12-fold respectively in
comparison to 20b. A compound with bulkier group 3-hydroxybutane at 6- position showed 4-fold
less potency than 20b. Similarly, tert-butyl group at 4- position reduced the potency by 5-fold (cf.
20d vs. 20b), or 6-position reduced by 13-fold (cf. 20n vs. 20e). A hydroxy group (20i) at 6-
position eliminated the potency of the molecule (K > 1000 nM) (Table 3). Incorporation of two
B
methyl groups as in 20o did not have additive impact on the potency (cf. 20o vs. 20b or 20e). We
also tested the 4-pyridyl derivative 20p, which showed complete loss of potency, whereas the 3-
pyridyl derivative 20q regained the potency and shown nearly same potency as 20a. Addition of
two methyl groups on 3-pyridyl ring (20r) similar to 20o indicated 10-fold loss of potency, thus
we have not subjected 3-pyridyl ring for further modification.
Table 3. Middle ring and right side ring modified EP2 antagonists. Potency and aqueous
solubilitya
16
ACS Paragon Plus Environment

Page 17 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Comp
ound
ID
EP2
Aqueousb
Solubility
(µM)
K
B
R =
(nM)
2
0a
20b
0c
0d
10.7
8.0
27
25
15
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
6.4
2
44.5
6.1
96
5
20e
20f
25
41
2
0g
0h
0i
20j
230
7.8
13
15
10
2
2
1220
32.5
5
2
0k
104
166
39
ND
ND
6
20l
2
0m
20n
79.3
9.7
25
15
2
0o
2
0p
20q
0r
>1000
8.3
88.1
38
17
100
2
aSchild K
2 measurements run in duplicate. ND = not determined The solubility of the compounds is
values are calculated similarly as indicated at Table 1 legend. K values are average of
B
B
b
4
6
measured in PBS buffer (pH 7.4) with 1% DMSO by nephelometry.
We then synthesized a new set of derivatives (Scheme 4) by modifying left side indole moiety and
the two carbon linker, keeping the middle ring as pyrimidine and the right side ring as 4,6-dimethyl
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
pyridyl ring as constant. As shown in Table 4, a fluoro, chloro, methoxy, difluoro or dichloro
derivatives 24a-e all retained high EP2 potency, whereas changing indole ring to pyrazolo-pyridine
ring as in 24g reduced the potency by 32-fold in comparison to 20o. Moreover, substituting the
methyl group at second position of indole with a trifluoromethyl group (24f) also reduced the
potency by 8-fold. Modification to the linker with ketone group next to indole (24h) eliminated
the potency, where as a hydroxy group as in 24i recued the potency 23-fold in comparison to 20o
suggesting unsubstituted ethylene linker is optimal for high EP2 potency within the scaffold.
Overall, the SAR study on the leads indicated that a number of compounds exhibit high EP2
potency.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 4. Modifications at indole ring and linker regions of the scaffold. Potency and aqueous
solubilitya
Entry
EP2
Aqueousb
Solubility
(µM)
Aqueous
solubility of
corresponding
HCl salt in
Water
solubility of
corresponding
HCl saltd
K
B
(nM)
c
R =
SGF (µM)
(mM)
2
0o
24a
4b
24c
4d
4e
9.7
13.0
28.3
20
150
5
>250
2.52
1.16
ND
ND
2
20
ND
ND
ND
20
>100
ND
1.73
2.38
ND
2
25.1
8.8
2
ND
2
4f
66
45
70
ND
ND
0.7
2
4g
254
ND
18
ACS Paragon Plus Environment

Page 19 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
4h
4i
>1000
183
ND
57
ND
ND
ND
ND
2
aSchild K
values are calculated similarly as indicated at Table 1 legend. K
2 measurements run in duplicate, except compound 20o for which the values are average of 6
repeats The solubility of the compounds is measured in PBS buffer (pH 7.4) with 1% DMSO by
nephelometry. The solubility of compounds measured in simulated gastric fluid at pH 2.0 by
nephelometry. Compound 24d is a milky-white solution in DMSO and 24 h is an inactive
compound. Therefore, their solubility is not determined by nephelometry, however, 24d.HCl is
values are average of
B
B
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
b
4
6 c
d
highly water soluble. Water solubility is determined by 24 h shake flask thermodynamic solubility
method in neat water. Briefly, the compounds were dissolved in water at 2 mg/mL and shaken at
2200 rpm for 24 h. Then, the solutions were filtered through 0.2 µ syringe filters and analyzed the
area under the curve (AUC) using HPLC, along with standard DMSO solution of each compound.
ND = not determined.
Table 5. Mouse liver microsomal (MLM) stability of selected novel EP2 antagonistsa
Compd. MLM T1/2
(minutes)
3a
8c
<2
2.9
13
2.9
6.5
8
17
12
96
16
8
d
8
g
1
4a
0b
20c
0e
0f
0h
0o
24a
2
2
2
2
2
17
17
26
15
2
2
4c
4e
2
4i
56
aThe compounds at 1µM were incubated in 0.125 mg/mL liver microsomes that are activated with
addition of NADPH (final concentration 1 mM) and quenched the reaction at 0, 5, 15, 30, 60 and
20 min using ice-cold acetonitrile. The % compound remaining was measured using LC-MS/MS
1
from which half-life was calculated. The work was done at CRO laboratories.
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Determination of aqueous solubility of EP2 compounds
Aqueous solubility is an important property that plays crucial role in success of a drug candidate.47
We tested several compounds for aqueous solubility by using two different methods. First, we used
a kinetic assay in which we dissolved the compounds in DMSO then made serial dilutions with
phosphate buffered solution (PBS) at pH 7.4, maintaining 1% DMSO in the solutions. Solubility
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
6
was determined by nephelometry. This assay works on the principle that when visible light is
passed through a solution, part of the incident radiant energy will be scattered. The intensity of the
scattered light is a function of the concentration of the dispersed phase. The nephelometric assay
is used to determine either the point at which a solute begins to precipitate out of a true solution to
form a suspension or the concentration at which a suspension becomes a solution when diluted
further. The scattered light will remain at a constant intensity until precipitation occurs, at which
point it will increase sharply. Several compounds shown in Table 2 and 3 displayed moderate
solubility (< 100 µM) in PBS at pH 7.4. However, interestingly, several of those compounds
including 8a-c, 8e-g, 14a, 20o, 24c have shown >100 µM aqueous solubility when determined in
simulated gastric fluid (pH 2.0), suggesting that the intrinsic basic nitrogen can be explored to
create salts that further increase solubility and facilitate the formulation. On this notion,
compounds 20o, 24a, c, d and 24f have been converted to hydrochloride salts by stirring them in
dichloromethane using dilute hydrochloride solution.
Second, to ask whether the hydrochloride salt compounds (20o, 24a, c d and 24f) are soluble in
48
water, we used a thermodynamic assay (shake flask method) where the compounds are shaken
(2 mg/mL, at 2200 rpm) in neat water and the soluble fraction was measured against standard
20
ACS Paragon Plus Environment

Page 21 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
DMSO solutions by HPLC or LC-MS/MS quantitation methods. As shown in Table 4, the
hydrochloride salt compounds displayed millimolar level aqueous solubility in water. It is worth
mentioning that the salt forms did not show increased solubility in PBS buffer at pH 7.4 by
nephelometry method in comparison to their neutral compounds. This is likely due to a drop in pH
in neat water, in comparison to a buffer. Indeed, the water pH was reduced from 6.4 to 3.5 when
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
0o.HCl was dissolved at 2 mg/mL. This observation is in line with our findings that the neutral
compounds have higher solubility in simulated gastric fluid (pH 2.0) than in PBS at pH 7.4.
Overall, the hydrochloride salts of compounds 20o, 24a, 24c and 24d showed 2.52, 1.16, 1.73 and
2
.38 mM solubility in neat water (Table 4).
Selectivity assessment of novel EP2 antagonists
The structural identity among the prostanoid receptor family is rather low. EP1, EP2, EP3 and EP4
4
9
share only 20-30% structural homology. In contrast, EP2 is more homologous to DP1 (44%) and
4
9
IP receptors (40%). In terms of cellular signaling, EP2, EP4, DP1 and IP induce cAMP mediated
2
+
cell signaling, whereas EP1 promotes Ca mediated signaling and EP3 inhibits cAMP mediated
signaling. Functionally, EP2 and DP1 receptors seem to promote inflammation in a variety of
disease conditions, whereas EP4 seems to act as pro and anti-inflammatory, and the IP receptor
1
5
seems to have cardioprotective role. Thus, we determined the selectivity of several potent EP2
antagonists (K < 50 nM for EP2) against DP1 first, and if any compound showed >100-fold
B
selectivity to DP1, then it was tested against EP4 and IP receptors. As shown in Table 6,
compounds 8a-b, 8e, 11a-b showed < 100-fold selectivity against the DP1 receptor, thus they were
not tested against EP4 and IP receptors. The compounds 8c, 14a, 20a, 20c, 20e, 20h, 20o, and 24a-
c displayed excellent selectivity against DP1, and subsequently against EP4 and IP receptors.
However, compounds 8g and 24c displayed < 200-fold selectivity against at least one of the three
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
DP1, EP4, and IP) receptors suggesting selectivity will depend on the individual structure of the
molecule within the scaffold. It is worth to note that we tested several compounds in the class for
cytotoxicity in the C6-glioma cell line. As exemplified in Table 6, selected compounds including
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
0o displayed no cytotoxicity until 50 µM indicating > 2100-fold in vitro therapeutic index
), except compound 8c which showed CC50 21 µM with a therapeutic index of 2100-
(CC50/EP
2
K
B
fold.
Table 6. EP2 Potency, selectivity index (S.I), and cytotoxicity (CC50) of selected novel EP2
antagonistsa
Entry
EP2
DP1
S.I.
(DP1/
EP2)
EP4
S.I
(EP4/
EP2)
IP
S.I
(IP/
EP2)
CC50
(µM)
K
B
K
B
K
B
K
B
(nM)
(µM)
(µM)
(
µM)
ND
ND
8
8b
8
8
8
a
23.2
9.5
10
48.4
2.9
48.6
33.6
29.6
10.7
6.4
6.1
7.8
9.7
8.3
13.0
28.3
20.0
1.0
0.9
3.0
0.8
1.2
1.7
0.5
6.6
>10
3.2
2.25
>10
7.32
1.77
9.0
42
95
300
16
410
35
15
300
>900
500
360
ND
ND
2280
ND
ND
ND
3.47
ND
0.73
ND
ND
21.5
>10
>10
>10
>10
>10
>10
>10
>10
>10
ND
ND
350
ND
25
ND
ND
730
>900
>1500
>1500
>1280
>1000
>1200
770
ND
ND
21
c
e
g
22.8
ND
ND
>50
ND
ND
>50
>50
>50
>50
>50
>50
>50
>50
>50
ND
3.91 1350
11a
11b
ND
ND
15.3
6.7
>10
>10
ND
ND
517
1
2
4a
0a
630
2
2
0c
0e
>1500
>1500
860
550
>1200
240
20h
20o
>1280 6.7
750
213
692
290
410
5.3
>10
3.1
>10
3.8
2
0q
4a
4b
24c
2
2
8.2
8.2
>350
193
>350
>500
aEP2 K
and IP K
nanomolar scale, whereas K
values are from average of 2 independent experiments run duplicate, except for 20o. EP4, DP1
values are from 2 independent experiments run in duplicate. K values for EP2 are in the
B
B
B
B
values for other receptors are shown in micromolar scale. CC50
22
ACS Paragon Plus Environment

Page 23 of 65
Journal of Medicinal Chemistry
(
concentration required to kill 50% C6glioma cells) values are from 1-2 experiments run in triplicate using
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
internal standard doxorubucin, which showed CC50 = 0.9 µM.
DMPK properties of novel EP2 antagonists
We tested several derivatives in pooled liver microsomal fractions to determine the half-life and
intrinsic clearance by the mouse liver, and to project in vivo pharmacokinetics. The starting compound
for the current study, 2a was metabolized quickly in liver microsomes (Table 5). To understand the
potential metabolites, we incubated compound 2a in mouse liver microsomes and investigated
metabolites after 5 minutes. As shown in Figure 2, the major metabolite is the amide bond cleaved
product D (MW 190 Da), which can be formed via intermediate A that would be further cleaved to
generate fragments B/C (MW 158 Da) and D. The compound D can also be generated via a vinylamine
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
F (MW 362 Da) intermediate, formation of which can be envisioned via hydroxylation at the CH
2
unit
(
E) next to 3-indole ring. Therefore, we synthesized compounds blocking the CH site with a hydroxy
2
group (e.g. 24i). The half-life of 24i in liver microsomes increased by 3-fold (t1/2, 57 min) compared to
its equivalent 20o, which showed 17 minutes of half-life (Table 5). However, the hydroxylated
derivatives displayed reduced EP2 potency compared with the equivalent compounds. (cf. 24i displayed
about 19-fold less potency than 20o, Table 4). In a similar experiment, the tetrazole compound 3 did
not produce any fragments even after 20 minutes of incubation in liver microsomes, suggesting the
metabolic hydroxylation or cleavage also depends on the right side moiety (Figure 2). The mouse liver
microsomal half-life data presented in Table 5 also indicate that a fluorine atom in place of
metabolically prone methyl group enhances the stability. For example, a fluoro- derivative 20c has 2-
fold higher half-life in comparison to 20b, like wise 20f has 8-fold higher half-life than 20e.
Regrettably, 20c is unstable in vivo, and 20f displayed less potency to EP2 receptor (Table 3), therefore,
these compounds are not promoted for further studies.
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. Metabolites identified from the lead EP2 antagonists in mouse liver microsomes.Compound
2a was incubated for 5 min and metabolic fragments (A-F) were detected using LC-MS/MS. Unlike
compound 2a, compound 3 has not shown any metabolic fragments even after 20 min of incubation
in mouse liver microsomes.
We estimated the CNS drug-like properties for several of these compounds before synthesis using the
CNS-MPO tool.43, 50 MPO scores indicated that several EP2 antagonists in the class as exemplified in
Table 7, display a desired MPO score of > 4, but several others do not achieve the desired score > 4
derived using their physicochemical properties. Compounds > 4 MPO score should have desirable
CNS activity and permeability features. However, the compounds 3 while it displayed > 5 score does
not have good permeability properties and does not get in to the brain (Table 7). A corollary to this, the
current lead compound 20o.HCl showed < 4 MPO score, yet crosses the blood-brain-barrier with in
vivo brain-to-plasma ratio 0.4. Likewise, compounds 8c, 8g and 14a behaved similarly. Overall, we
have seen a positive correlation between MPO score to in vivo brain to plasma ratio. Nonetheless, the
24
ACS Paragon Plus Environment

Page 25 of 65
Journal of Medicinal Chemistry
1
2
3
MPO score does not quantitatively predict the brain permeability. We tested several compounds for
permeability across MDR1 expressed MDCK cell line to investigate the potential of blood brain barrier
(BBB) permeability within the class. As shown in Table 7, we found many of these compounds have
good passive permeability from the apical-to-basolateral (A-B) side as well as the basolateral-to-apical
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
-6
side (> 0.6 x10 cm/s) in comparison to compound 3. and their efflux ratio is < 3, except for compound
4a which displayed ER ratio 16 indicating it may be a substrate for efflux pumps (Table 7). On the
1
other hand, the compound 20o found to show good permeability and showed similar efflux ratio in the
presence and absence of an efflux pump inhibitor verapamil, confirming that it is not the substrate of
efflux pumps. (Table 7). Moreover, compound 20o showed 0.4 % plasma protein unbound fraction in
mouse plasma proteins, but we were not able to determine its fraction unbound in the relevant brain
compartment (i.e., interstitial fluid) due lack of efficient method to extract cerebrospinal fluid (CSF)
from mouse brain.
Table 7. BBB-permeability, mouse and human liver microsomal stability, key pharmacokinetic
properties for selected compoundsa
Entry MLMt HLM
In vivo
Brain- Permeability across
Efflux
ratio
(B-A /
A-B)
MPO score
d
1
/2(min)
t
1/2
plasma t
to-
plasma
MDR1-MDCK cell
a
c
(min)
1/2 (h)
line
(Desired
score for
CNS
ratiob
permeability
>
4.7
4)
10
11
1.7 (ip/po,
1.7
B-A: 27.1 x 10-6
cm/s
1.1
0.3
1
3
1
0 mg/kg
dose)
A-B: 25.4 x 10-6
cm/s
-6
>60
>60
> 6h (po,
0mg/kg)
0.02
B-A: 0.12 x 10
5.0
1
cm/s
A-B: 0.41 x 10-6
cm/s
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
8
c
2.9
2.9
6.5
17
ND
ND
ND
87.7
<1h (ip,
0.5
ND
0.16
0.4
B-A: 51.0 x 10-6
2.6
2.7
16
3.8
3.7
4.3
3.6
1
0mg/kg)
cm/s
A-B: 19.8 x 10-6
cm/s
8g
< 1h (ip,
B-A: 55.3 x 10-6
cm/s
1
0mg/kg)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
-6
A-B: 20.6 x 10
cm/s
B-A: 71.1 x 10-6
cm/s
1
4a
0o
< 1h (ip,
1
0mg/kg)
A-B: 4.54 x 10-6
cm/s
-6
2
1.1 h (IP,
B-A: 21.5 x 10
2.1
1
0 mg/kg)
&
cm/s
A-B: 10.2 x 10-6
cm/s
2
.4 h (po,
50 mg/kg)
In the presence of a pgp-inhibitor (verapamil)
B-A: 22.2 x 10-6
cm/s
1.7
2
0o
A-B: 13.2 x 10-6
cm/s
a
MLM = mouse liver microsomes; HLM = human liver microsomes. In liver microsomal stability tests,
1 µM compound test compound was incubated with 0.5 mg/mL liver microsomes for compounds 1 and
3. However, other compounds in the table were incubated with 0.125 mg/mL liver microsomal
concentration. The plasma half-life is estimated based on the 3-time point B/P ratio studies conducted
with single injection of 10 mg/kg dose to male mice. A full-scale pharmacokinetic study with 8-time
points was done for selected compounds (1, 3, 20o). Brain-to-plasma are derived from peak
concentrations observed at 0.5 h after injection in to mice. BBB potential was determined using
b
c
MDR1-expressed cell monolayers. All these studies are conducted at CRO laboratories following
industry standard procedures. See Table 1 for the physicochemical properties used to calculate the
d
MPO score.
To determine in vivo brain-to-plasma ratio in mice, we tested several compounds whose mouse liver
microsomal stability was greater than 15 min, by administering a dose of 10 mg/kg via intraperitoneal
injection. Analysis of the compound concentration in the plasma and the brain tissues at 0.5, 2, and 4 h
time points suggested that most of these compounds, exemplified as in Table 7 for 8c, 8g & 14a, will
have a plasma half-life below 1 h, except compound 20o. A subsequent pharmacokinetic analysis on
26
ACS Paragon Plus Environment

Page 27 of 65
Journal of Medicinal Chemistry
1
2
3
2
0o.HCl with concentration analysis at 8-time points after a single i.p. injection (10 mg/kg) (Figure 3)
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
indicated that it has terminal plasma half-life 1.1 h with clearance rate of 124.1 mL/min/kg and brain-
to-plasma ratio of 0.4. Additional studies with oral gavage dosing (50 mg/kg, B.I.D. dosing 8h apart)
indicated that the plasma half-life could be extended to 2.4 h for 20o.HCl.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
.
Figure 3. Pharmacokinetic parameters of 20o.HCl Compound 20o.HCl was administered in to male
C57BL/6 mice via intraperitoneal injection at single dose of 10 mg/kg in the vehicle of 5% NMP, 5%
Solutol-HS15 and 90% water. Concentrations in plasma and brain are determined by LC-MS/MS and
plotted against time.
Competitive mode of EP2 antagonism by 20o.HCl and derivatives.
To determine whether the novel derivatives in this class exhibit competitive antagonism of EP2, we
tested several compounds in a concentration-response manner against PGE
EP2 receptors. The Schild K indicates the antagonist concentration required for a two-fold rightward
shift in the PGE concentration-response curve. Schild K values are derived by the equation log (dr −
– log K , where dr = dose ratio, i.e. the fold shift in agonist EC50 caused by the antagonist,
is antagonist concentration. As illustrated in Figure 4B, a linear regression of log (dr-1) on log X
2
concentration effect on
B
2
B
1
) = log X
B
B
X
B
B
with slope of unity characterizes a competitive antagonism. A smaller K value indicates a higher
B
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
inhibitory potency. As shown in Figure 4A and C, compound 20o (TG11-77 neutral and the
hydrochloride salt form) induced a concentration-dependent, parallel rightward shift in the PGE
concentration-response curve. Schild regression analyses (Figure 4B) is consistent with a competitive
mechanism of antagonism on EP2 with average Schild K 9.7 nM and average slope value 1.05. The
average K value is used throughout in this manuscript for discussion and comparisons. Moreover,
several other compounds synthesized in this class displayed a concentration-dependent rightward shift
of PGE EC50 and with slope of unity (SI Figure 2). Thus, the mechanism is competitive in general for
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
B
B
2
the class of EP2 antagonists presented in this study. However, compound 20o did not show a
concentration-dependent inhibition of DP1 receptor, indicating it is selective for EP2 over DP1 with
selective index (DP1 K
B
/EP2 K ) of 750-fold. Moreover, this compound also showed high selectivity
B
against EP4 and IP receptors with selective index of 550-fold and >1000-fold respectively (Table 6).
28
ACS Paragon Plus Environment

Page 29 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. Competitive antagonism of EP2 receptor by 20o. A: Compound 20o (TG11-77 neutral)
inhibited PGE
Schild regression analysis is performed to determine the modality of antagonism by this compound.
C. Hydrochloride salt of 20o (20o.HCl) similarly inhibited PGE -induced human EP2 receptor
activation in a concentration-dependent manner. Schild K values for neutral compound and the
2
-induced human EP2 receptor activation in a concentration dependent manner. B.
2
B
hydrochloride salt along with their slope values are shown in inset of Figure 4B. D. Concentration-
response test of 20o.HCl on DP1 receptors indicates it does not significantly inhibit DP1 receptor
activation by agonist BW245C. Data were normalized as percentage of maximum response; points
represent mean ± SEM (n = 3).
EP2 antagonists display anti-inflammatory properties in a novel microglia cell line expressing
human EP2.
29
ACS Paragon Plus Environment